Arcing prevention during post-repair EHT testing

[3Phase615V]

Hi,

Would anyone know if there is a safe method of testing an (oscilloscope) EHT module following de-potting and EHT winding repair, so as to prevent arcing between the multiplier's caps and diodes and perhaps the transformer legs? This testing would be before finally re-potting the EHT module PCB back inside its case. The EHT output voltage should be around 13kV when running and the gaps between the legs of the high voltage caps and diodes is as little as 2mm in places.

Thanks

[Radio Wrangler]

The potting compound has a higher volts/mm capability than air. It allows multipliers to be made smaller than if they were open-built. So it won't go to full voltage until potted.

So you're limited to testing on reduced voltage until you take the plunge and re-pot.

David

[3Phase615V]

Originally Posted by Radio Wrangler

So you're limited to testing on reduced voltage until you take the plunge and re-pot.

Hi and thanks for your reply, RR.

Do you mean that the EHT voltage will be automatically reduced in air or that I need to reduce the voltage myself in order to carry out testing of the EHT module in air?

Thanks

[Radio Wrangler]

If you try testing in air, the components are so close together that you may get arcing and resulting damage.

Some older valved audio generators like the HP200 series could make you several tens of volts at tens of kilohertz, and you could do a basic function test to make sure a multiplier was rectifying and multiplying properly.

After that you had to pot them up and let them cure before trying them at full voltage.

I worked on a number of instruments with CRTs in them and careful measures were needed to make the EHT reliable. Evenat the 2kV level for cathode/grid circuits having large boards with routed slots was necessary. Multipliers for PDAs used to be huge if not potted, and reliability suffered from dust and general contamination attracted in.

David

[3Phase615V]

So it looks like reducing the voltage is my best and perhaps only option, but it won't allow me to test the EHT module at full tilt until I re-pot the thing. I don't have access to a HP200-type generator so I'll have to make do with using what I have and can do.

Thanks for your advice - much appreciated.

[Keith]

I remember seeing aerospace HVPSU (28kV@2mA) prototypes being tested under some clear liquid. Sadly I can't recall what it was. We also used to make 15kV units in sealed containers filled with SF6 gas but I suspect that is definitely a H&S no-no these days!

[Jac]

Keith,

The liquid very likely was FC-43 "fluorinert electronic liquid" made by 3M.

3Phase615V,

As a general rule of thumb, you can use the fact that arcing occurs at a field strength of 1 kV per mm in air.
So for 13 kV you'd need about at least half an inch of distance.
A no-go for testing without potting (which originally might have been done in a vacuum, to avoid air pockets).

As a means to be able to test it, you could immerse the unit in clear vegetable oil. Do your best by vibrating/shaking to remove as much air as you possibly can.
Make sure you remove all oil/grease before potting it.

Hope this helps.

Jac

[Ed_Dinning]

Hi, in the old days this type of test was often done in the EHV lab in a "bucket" of SF6, which being heavier than air, stayed put.
This is not a toxic gas, but very severely ozone depleting, so covered by all sorts of regs.
It is still used, under controlled conditions on a lot of HV switchgear to this day.

As has been suggested, you can use clear veg oil, but it should be heated first the remove dissolved air. Beter if you can get it are the minerals oils used in switchgear and cables. Treat in a similar way to remove air and moisture.

Ed

[3Phase615V]

Excellent stuff! Thank you very much for all your respective pieces of advice. It did occur to me to investigate using some sort of oil, but I am not keen on having a greasy mess to deal with and the oil potentially lifting the polyester transformer insulation tape from my very neat EHT winding. I was also thinking about glue gun glue or silicon sealant as it's pretty easy to remove once dry, that is assuming I can get bubble-free seals over the exposed metallic parts carrying high voltage. I have read of successful cases, at least anecdotally, of individuals using standard glue gun glue to insulate up to 22kV contact points.

[3Phase615V]

Hi Jac,

I read via the link below that the spark/arc gap formula is Gap in CM x 30,000, which in my case would mean 6kV before an arc might occur across a 2mm gap. Unfortunately the article does not explain where the figure of 30,000 comes from but I would imagine your calculation is probably more real life and from experience!

https://sciencing.com/calculate-volt...s-8776030.html

[Radio Wrangler]

Originally Posted by Jac

As a general rule of thumb, you can use the fact that arcing occurs at a field strength of 1 kV per mm in air.

That's at sea level, which is OK for London.

It's a bit lower in Colorado and other elevated parts of the world so watch out for them.

Some of us have to handle altitudes of over 60,000 ft with our designs!

David

St Elmo's fire? No, St Aidan's - I was running less than 5 Watts. (A QRP joke)

[Jac]

Absolutely true David.

Air humidity also plays a role and let's not forget creep currents.

Jac

[3Phase615V]

Ah I see - thank for clarifying.

May I ask what are your thoughts are on my previous message?

[GrimJosef]

What really matters is the 'field enhancement factor' that sharp edges bring. Although the average electric field across a 5mm gap with 5kV across it is 1kV/mm, if one of the electrodes has a sharp edge the field will be distorted there and can be raised several times. A factor of 5 is straightforwardly achievable, so the local peak field could be 5kV/mm, or more. This will cause the air to break down locally, forming a plasma blob. The resulting supply of ions and hard ultraviolet radiation will cause the rest of the gap to break down in no time. In a previous life I used to design and build very high voltage equipment (megavolts) including spark-gap switches, both triggered and self-breaking.

If this piece of kit is designed to work potted then apart from fairly trivial tests at low voltage there really is no way of checking it properly other than by potting it. SF6 can help a bit. Oil is better, but is still not the same as potting compound. The pourable silicones e.g. https://uk.farnell.com/acc-silicones...ite/dp/2387242 are also good (but not cheap) although as others have said you do have to take care with bubbles.

Cheers,

GJ